INDEX

Explanations

words related to editors or editing

oai_token-act-pair · gpt-3.5-turbo

mentions of editors in various contexts

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

NetMessage

-0.81

asons

-0.66

yrs

-0.65

nants

-0.65

ramid

-0.65

priority

-0.62

 bonding

-0.61

ulus

-0.61

ples

-0.61

á

-0.61

POSITIVE LOGITS

 editor

3.85

 Editor

3.07

 editors

2.89

editor

2.81

 Editors

2.24

Editor

2.06

 editing

1.85

 editorial

1.83

 publisher

1.83

 Editorial

1.67

Activations Density 0.016%

words related to editors or editing

mentions of editors in various contexts

No Comments

No Known Activations